New Evidence on the Effect of

Changes in College Admissions

Policies After the Pandemic
Examining the Impact of Test-Optional and Race-Neutral
Policies on College Applications, Admissions, Enrollment,
and First-Year Outcomes

Michael Bloem
Ashley Edwards
Parker Goyer
Jessica Howell
Michael Hurwitz
Samuel Imlay
Jennifer Ma

College Board Research

June 2025

1
 Executive Summary
The Admissions Research Consortium (ARC) is a collaborative research initiative that
provides participating institutions insight into their own and their peers’ admissions
processes and outcomes, as well as changes in student behavior influenced by the covid-
19 pandemic. Previously published consortium research examines changes in fall 2021
and fall 2022 applications, admissions, and enrollment, with a focus on students’ test
score disclosure choices (Howell et al., 2022; Goyer et al., 2023) and first-year college
grades, credit accumulation, and retention into the second year for fall 2021 first-year
enrollees compared to prior cohorts of first-year students at ARC institutions (Edwards et
al., 2023). New research from the consortium presents updated data on trends through
the fall 2024 college application cycle and analyses of college student outcomes through
the 2023-2024 academic year. The time period studied includes notable changes to the
college admissions landscape, including widespread test-optional policies beginning with
the fall 2021 entering class and race-neutral policies beginning in fall 2024. The results
inform discussions about future admission policies and practices as well as efforts to
support student success.

Approximately 60 colleges, representing selective public and private four-year institutions

in the U.S., provided College Board with data on their applications, admissions,
enrollment from the fall 2018 through the fall 2024 entering cohorts, and performance,
credits, and retention from the 2018-19 to 2023-24 academic years. Four themes
emerged from the analyses of the most contemporaneous data.

Theme 1: Changes to colleges’ funnels largely reflect more applications, more

offers of admission, and increased four-year college enrollment.

Between fall 2020 and fall 2024, applications to consortium institutions increased by
roughly a half of a million applications, or a 38% increase. Application growth in the past
four cycles is strong among all student subgroups and across all institution types, with
stronger growth in applications to public than private institutions. The number of students
admitted by consortium institutions increased by 20% between fall 2020 and fall 2024,
driven predominantly by admission growth of 24-30% at public institutions. By contrast,
More Selective Private institutions made 12% fewer offers of admission in fall 2024 than
in the years prior to the pandemic. The only applicant subgroup to receive fewer offers of
admission in fall 2024 than before the pandemic are those students with high school
GPAs (HSGPAs) in the bottom three quintiles. Finally, fall 2024 enrollment in consortium
institutions is 11% higher than in fall 2020, balanced by more than 20% growth at
Selective Public institutions and no enrollment growth at More Selective Private
institutions. Enrollment growth increased for all student subgroups except those in the
bottom three HSGPA quintiles and among White students, for whom enrollment growth
was flat.

2
 Theme 2: Across all consortium institutions, student body composition by
race/ethnicity and socioeconomic status is largely unaffected by widespread test-
optional (in fall 2021) or by policy changes to the consideration of race (in fall
2024).

Small, annual increases in the racial/ethnic composition of applicants, admitted students,

and enrollees at consortium institutions mimic patterns of racial diversification observable
among high school graduates in the U.S. (Lane et al., 2024). Widespread test-optional
admissions policies in fall 2021 were associated with a 3.8 percentage point increase in
the share of enrollees who are Black, Hispanic, Native, or Two or More Races at More
Selective Private institutions. There were negligible changes in these underrepresented
minority student shares in other institutional segments. That increase at More Selective
Private institutions is partially offset by a 2.8 percentage point decrease in
underrepresented minority student shares at More Selective Private institutions in fall
2024 when the Supreme Court ruling changed admissions policy and practice related to
applicant race/ethnicity.

The socioeconomic composition of students at consortium institutions is unchanged

between 2018 and 2024 with roughly 25% of enrolled students coming from higher-
challenge neighborhoods.

Theme 3: Roughly half of applicants to consortium colleges choose to disclose a

test score for consideration in admissions, a choice that is most strongly related to
their test score (relative to the college where they are applying) and very little else.

Among applicants with the same test score applying to the same college, the probability
that a student chooses to disclose a test score is the same for first generation and non-
first generation students, very similar across different racial/ethnic student subgroups, and
the same for STEM and non-STEM majors. Applicants with lower high school GPAs
(HSGPA) are more likely to disclose test scores than otherwise similar students with
higher HSGPAs. Over the test-optional cycles from fall 2021 to fall 2024, test score
disclosure rates declined slightly from 52% to 48% of applications.

Theme 4: For both score disclosers and non-disclosers, test scores are strong
predictors of first-year college outcomes like grades, credit accumulation, and
retention to the second year.

Applicants who disclose test scores are academically stronger (as measured by both high
school grades and test scores) than applicants who do not disclose scores, so it follows
that score disclosers have strong first-year outcomes—GPAs, credit accumulation, and
retention rates—than non-disclosers. The relationship between test scores and college
outcomes persists among students with the same high school grades. Students with
lower test scores, who are more likely to withhold them, have the highest risk of lower
first-year college grades, especially in STEM courses.

3
Table of Contents
Executive Summary ............................................................................................................ 2

The Admissions Research Consortium (ARC).................................................................. 8

ARC Data Sample and Definitions ........................................................................................ 8

Applications, Admissions, and Enrollments ................................................................... 11

Aggregate Enrollment Funnel Insights ................................................................................11

Applications ........................................................................................................................14

Admissions .........................................................................................................................16

Enrollments .........................................................................................................................18

Student Body Composition .............................................................................................. 20

Racial/Ethnic Composition ..................................................................................................20

Socioeconomic Composition ...............................................................................................23

Test Score Choices ........................................................................................................... 26

Score Disclosure, Withholding, and Absence ......................................................................26

Score Disclosure Model ......................................................................................................27

First-Year Outcomes ......................................................................................................... 34

Precollegiate Achievement .................................................................................................. 34

First-Year College Outcomes ..............................................................................................38

First-Year Grades ...............................................................................................................39

First-Year Academic Risk ....................................................................................................40

First-Year Credit Accumulation ...........................................................................................42

First-Year Retention ............................................................................................................ 44

Discussion of College Outcomes.........................................................................................46

Outcomes in STEM and Non-STEM Coursework ................................................................ 47

Conclusion ........................................................................................................................ 49

4
Appendix ........................................................................................................................... 50

References ........................................................................................................................ 59

About the College Board .................................................................................................. 61

List of Figures

Figure 1: ARC Colleges’ Enrollment Funnel, Fall 2018 to Fall 2024..................................... 11

Figure 2: Enrollment Funnel for ARC Colleges and Segments, Fall 2018 to Fall 2024 ........ 12
Figure 3: Percentage Change in Applications, Admissions, and Enrollment Between Fall 2020
and Fall 2024, ARC Colleges and Segments ..................................................................... 13
Figure 4: Percentage Change in Applications to ARC Colleges Between Fall 2020 and Fall
2024, by Student Characteristics ........................................................................................ 14
Figure 5: Percentage Change in Admission Offers at ARC Colleges Between Fall 2020 and Fall
2024, by Student Characteristics ........................................................................................ 16
Figure 6: Percentage Change in Enrollment at ARC Colleges Between Fall 2020 and Fall 2024,
by Student Characteristics .................................................................................................. 18
Figure 7: Racial/Ethnic Composition of Applicants from Fall 2018 to Fall 2024.................... 20
Figure 8: Racial/Ethnic Composition of Admitted Students from Fall 2018 to Fall 2024 ....... 21
Figure 9: Racial/Ethnic Composition of Enrolled Students from Fall 2018 to Fall 2024 ........ 22
Figure 10: Racial/Ethnic Composition of Enrolled Students from Fall 2018 to Fall 2024, by ARC
Segment ............................................................................................................................. 23
Figure 11: Socioeconomic Composition of Applicants from Fall 2018 to Fall 2024 .............. 24
Figure 12: Socioeconomic Composition of Admitted Students from Fall 2018 to Fall 2024 .. 24
Figure 13: Socioeconomic Composition of Enrolled Students from Fall 2018 to Fall 2024 .. 25
Figure 14: Socioeconomic Composition of Enrolled Students from Fall 2018 to Fall 2024, by
ARC Segment ..................................................................................................................... 26
Figure 15: Score Disclosure, Withholding, and Absence Among Fall 2021 to 2024 Applications,
by ARC Segments...............................................................................................................27

Figure 16: Predictors of Test Score Disclosure for Fall 2024 ...............................................28

Figure 17: Probability of Test Score Disclosure Among ARC College Applications, Fall 2021 to
2024.................................................................................................................................... 29

Figure 18: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by HSGPA........................................................................................................................... 30

5
Figure 19: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by Parental Education ......................................................................................................... 31
Figure 20: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by Race/Ethnicity ................................................................................................................ 32
Figure 21: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by College Major ................................................................................................................. 33

Figure 22: Distribution of High School Grades Among Enrolled First-Year Students at ARC
Institutions in Fall 2024, by Score Disclosure Status and ARC Segment ............................. 34
Figure 23: Distribution of High School Grades Among Enrolled First-Year Students at ARC
Institutions in Fall 2024 versus Fall 2020, by ARC Segment................................................ 35
Figure 24: Distribution of SAT Scores Among Enrolled First-Year Students at ARC Institutions in
Fall 2024, by Score Disclosure Status and ARC Segment .................................................. 36
Figure 25: Distribution of SAT Scores Among Enrolled First-Year Students at ARC Institutions in
Fall 2024 versus Fall 2020, by ARC Segment ..................................................................... 37
Figure 26: Average FYGPA by Test Score Band and Disclosure Status, 2023-24................ 39
Figure 27: Average FYGPA by Test Score Band, HSGPA, and Disclosure Status, 2023-24
........................................................................................................................................... 39

Figure 28: Institution-Specific Changes in Average FYGPA, 2023-24 versus 2018-29 ........ 40
Figure 29: Academic Risk by Test Score Band and Disclosure Status, 2023-24 .................. 41
Figure 30: Academic Risk by Test Score Band, HSGPA, and Disclosure Status, 2023-24 ... 41
Figure 31: Institution-Specific Changes in Academic Risk, 2023-24 versus 2018-19 .......... 42
Figure 32: Average Credits by Test Score Band and Disclosure Status, 2023-24 ................ 43
Figure 33: Average Credits by Test Score Band, HSGPA, and Disclosure Status, 2023-24
........................................................................................................................................... 43
Figure 34: Institution-Specific Changes in Average First-Year Credits, 2023-24 versus 2018-19
.......................................................................................................................................... 44
Figure 35: First-Year Retention by Test Score Band and Disclosure Status, 2023-24 .......... 45
Figure 36: First-Year Retention by Test Score Band, HSGPA, and Disclosure Status, 2023-24
........................................................................................................................................... 45
Figure 37: Institution-Specific Changes in Retention Rates, 2023-24 versus 2018-19 ........ 46
Figure 38: Average FYGPA at ARC Institutions by SAT Score Over Time............................ 47
Figure 39: Average FYGPA in STEM and Non-STEM College Coursework in Post-Pandemic
Cohorts, by SAT Score ........................................................................................................ 48

6
Figure A1: Average FYGPA at ARC Institutions by SAT Score Over Time, by ARC Segment
........................................................................................................................................... 54
Figure A2: Average FYGPA in STEM and Non-STEM College Coursework in Post-Pandemic
Cohorts, by SAT Math Score............................................................................................... 55
Figure A3: Average FYGPA by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment ............................................................................................................................. 56
Figure A4: Average FYGPA by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment ............................................................................................................................. 56
Figure A5: Average FYGPA by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment ............................................................................................................................. 57
Figure A6: Average FYGPA by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment ............................................................................................................................. 57

List of Tables

Table 1: Annual Percentage Changes in Applications to ARC Colleges Between Fall 2020
and Fall 2024, by Student Characteristics........................................................................ 15
Table 2: Annual Percentage Changes in Admission Offers Between Fall 2020 and Fall 2024,
by Student Characteristics................................................................................................ 17
Table 3: Annual Percentage Changes in Enrollment Between Fall 2020 and Fall 2024, by
Student Characteristics .................................................................................................... 19
Table 4: Average College Outcomes at ARC Institutions, by Academic Year ....................... 38
Table A1: Attributes of ARC and Non-ARC Institutions by Segment, Sample 1 .................... 50
Table A2: Attributes of ARC and Non-ARC Institutions by Segment, Sample 2 .................... 51
Table A3: Average First-Year College Outcomes by Score Disclosure, Withholding, and Absence
........................................................................................................................................... 52
Table A4: Average First-Year College Outcomes, by Academic Year and ARC Segment ..... 53

7
The Admissions Research Consortium (ARC)
The Admissions Research Consortium (ARC) is a collaborative research initiative that aims to
help participating institutions gain insight into their own and their peers’ admissions processes
and outcomes, as well as changes in student behavior influenced by the covid-19 pandemic.
Guided by a Core Advisory Committee and a Research Advisory Committee1, evidence and
insights from ARC inform future practice and policy in the years following the pandemic. This
research presents data on trends through the fall 2024 college application cycle and analyses of
college student outcomes through the 2023-2024 academic year. The time period studied
includes notable changes to the college admissions landscape, including widespread test-
optional policies beginning with the fall 2021 entering class and race-neutral policies beginning
in fall 2024.

ARC Data Sample and Definitions

In 2024, ARC institutions shared administrative data on applications, admissions, and

enrollment from fall 2018 to fall 2024, as well as performance, credits, and retention from the
2018-19 to 2023-24 academic years. These data, which also include a robust set of student
demographic and academic variables, were merged with College Board assessment data to
enable insight into students who disclosed and withheld SAT test scores under widespread test-
optional admissions policies brought on by the pandemic. We provide evidence on student and
college choices based on the combined dataset, which covers a sample of 60 selective four-
year public and private nonprofit institutions over seven application cycles. We categorize the 60
ARC institutions into four institutional segments defined by institutional control and selectivity:2,3

• More Selective Private Colleges: 21 private institutions with admit rates below 25%
• Selective Private Colleges: 18 private institutions with admit rates above 25%
• More Selective Public Colleges: 13 public institutions with admit rates below 60%
• Selective Public Colleges: 8 public institutions with admit rates above 60%

ARC colleges in the More Selective Private and More Selective Public segments are quite
representative of all institutions in those same segments, while ARC colleges in the Selective
Private and Selective Public segments tend to be more selective than non-ARC institutions in
those same segments (see Appendix Tables A1 and A2 for more detail on differences).

1
The ARC Core Advisory Committee consists of representatives from the Association for Institutional Research [AIR],
American Association of Collegiate Registrars and Admissions Officers [AACRAO], National Student Clearinghouse
[NSC], and senior admission and enrollment practitioners. The Research Committee consists of academic and
institutional researchers.
2
Admit rates are based on data from ARC colleges and are calculated as a weighted average using Fall 2018 to Fall
2020 application data.
3
There are 56 colleges for which we have data on first-year GPA, credits earned, and retention. See Appendix Table
A2 for more detail on this subset of institutions.

8
This research employs the following terms and definitions:

Application cohorts, application cycles, and admission cycles are indexed according to the
fall entry term for which students applied (e.g., fall 2024). Because ARC studies fall admissions
processes and their outcomes, all analyses in this report group students according to the fall
entry term for which they applied, even if students deferred first-year enrollment to a later entry
term. Thus, applicants for fall 2023 who deferred enrollment to fall 2024 are grouped as
enrollees from the fall 2023 application cohort.

Applications/Applicants refer to the aggregate number of applications ARC institutions

collectively received and evaluated, which is notably larger than the number of unique
applicants who applied to ARC institutions, since some students applied to several ARC
institutions.

Admissions/Admits/Admitted Students refers to the aggregate number of admission offers

ARC colleges collectively extended, while admits and admitted students refer to unique students
offered admission, and admit rate refers to the fraction of applications offered admission.

Data from ARC institutions were merged with College Board assessment data to enable
research insights into test score disclosure. Because the final dataset includes all observable
test scores from either the institution or College Board, when considering SAT/ACT scores, we
distinguish between disclosed SAT/ACT scores, withheld SAT scores, and students with no test
score or a withheld ACT score. Disclosed SAT/ACT scores are SAT/ACT scores that students
submitted to ARC colleges for consideration in the admissions process. Withheld SAT scores
are SAT scores that applicants withheld from colleges, but that are observable in College Board
administrative data. We define withheld SAT scores from College Board administrative data by
the combination of a student’s highest SAT section scores. Finally, a third category of students,
referred to as No-Test / Withheld ACT scores, are students who did not disclose an SAT or
ACT score to the ARC institution, who do not have a recorded SAT score on file with College
Board, and who may either have an ACT score that they withheld or never took an SAT or ACT
assessment. In the figures below, we occasionally display data for all three of these categories
of students. More commonly, the latter two categories are combined and referred to as Non-
Disclosers to simplify the visual representation of the data.

Underrepresented Minority (URM) students are defined as students who are Black,
Hispanic/Latinx, Native American, Native Hawaiian/Pacific Islander, or Two or More Races.

Parental education data come from ARC colleges and are based on the institutions’ own ways
of collecting and designating parental education. In some analyses, we also use College Board
data on parental education from the Student Data Questionnaires accompanying assessments.
First generation is defined as not having a parent who has completed a bachelor’s degree.

Low-income status data come directly from ARC colleges, and reflect the institution’s own
definition of low-income.

9
Landscape neighborhood challenge is a measure of students’ neighborhood background, as
a proxy for students’ socioeconomic status. Landscape neighborhood challenge is a
neighborhood-level attribute constructed based on U.S. Census data and a nationally
representative sample of high school graduates, and it has been used by many participating
colleges in recent application cycles. Neighborhood challenge levels are expressed on a 1-100
percentile scale, where higher values indicate higher levels of challenge related to educational
opportunities and outcomes. For example, a neighborhood with a Landscape challenge level of
64 has a higher level of educational challenge than 64% of neighborhoods in the U.S. Likewise,
20% of U.S. neighborhoods fall into each quintile of Landscape neighborhood challenge.

Feeder high schools are defined as high schools that sent more than 30 applications to an
institution over the fall 2018-2020 application cycles.

Twenty-five participating colleges provided data on applicants’ recalculated high school grade
point averages (HSGPAs). Because different colleges employ different grade scales when
recalculating applicant HSGPAs, we created five HSGPA quintiles using college-specific
HSGPA quintile cut-points based on the HSGPA distribution among each college’s fall 2018-
2020 applicants. At each college reporting recalculated HSGPAs, roughly 20% of fall 2018-2020
applicants fall into each quintile of recalculated HSGPA. We also use self-reported HSGPA
data from College Board’s Student Data Questionnaire in analyses when HSPGA data are not
available from colleges. The correlation between actual and self-reported HSGPA is high,
ranging from the low-0.70s to the mid-0.80s (see Marini, Young, and Shaw (2021) for more
details).

First-Year GPA (FYGPA) are data provided by participating institutions on the average grades
earned by enrolled students in a particular academic year.

STEM and Non-STEM FYGPA is also college-provided data on the average grades earned in
particular types of college courses in the first year. STEM FYGPA captures course performance
in biological sciences and other health-related professions, computer science, engineering,
math, physical sciences, and other science and military technologies. Non-STEM FYGPA
includes all coursework outside of these domains. Students taking no STEM courses in their
first year have a blank value for STEM FYGPA and are not included in analyses examining
STEM performance. Similarly, students only taking STEM courses are not included in analyses
focused on Non-STEM course performance.

First-Year Credits are defined as the number of credits earned in the first year of college on the
institution-specific credit scale. In some analyses, these data are normalized to a 30-credit scale
to facilitate aggregation across ARC institutions and segments.

Retention is defined as a student re-enrolling at the same institution in their second year of
college.

10
Applications, Admissions, and Enrollments

Aggregate Enrollment Funnel Insights

Between the fall 2018 and fall 2024 admission cycles, ARC institutions received more than 10
million applications, extended more than 4 million offers of admission, and enrolled nearly 1.2
million first-year students. This sample comprises the dataset analyzed in this report.4
Figure 1 displays counts of applications, admission offers, and first-year enrollment at all ARC
institutions from fall 2018 through fall 2024. Applications sharply increased from fall 2020 to fall
2021, by 15%, and continued to increase, at a lower rate (on average 6.3%), from fall 2021 to
fall 2024. Admission offers steadily increased since fall 2021. Enrollment increased from fall
2020 to fall 2021 at a higher rate than previous or subsequent cycles.

Figure 1: ARC Colleges’ Enrollment Funnel, Fall 2018 to Fall 2024

4This analysis sample excludes about 1.7 million applications to ARC institutions that were incomplete or withdrawn
before an admission decision was reached.

11
Figure 2 compares funnel trends over seven years at all ARC institutions and the four
institutional segments. To better compare funnel trends across segments that differ in size and
selectivity, we express each data series as an index anchored to the fall 2018 entering cohort,
which corresponds to the start of ARC historical data. Each index has a value of 100 in the base
year.

Figure 2: Enrollment Funnel for ARC Colleges and Segments, Fall 2018 to Fall 2024

Between the fall 2018 and fall 2024 application cycles, total applications to ARC institutions
increased by 44%, but there is considerable variation across the four segments, with Selective
Public ARC institutions experiencing the largest application growth (60%) over these seven
cycles. Over this same period, admission offers from ARC colleges grew by 34%. Admissions
growth occurred in all institutional segments except More Selective Private colleges, where fall
2024 admission offers are 9% lower than in fall 2018.

Since fall 2018, yield rates at ARC institutions have declined by 17% overall. A similar decline
occurred in all institutional segments, except for the More Selective Private institutions, where
yield increased by 11% since fall 2018. Finally, between fall 2018 and fall 2024, enrollment at
ARC colleges increases by 11% overall. All institutional segments experience growth in

12
enrollment between fall 2020 and fall 2021, but Selective Public colleges experience the largest
growth—22%—between fall 2018 and fall 2024. Both private college segments experience
minimal enrollment growth over this period.

Figure 3 shows aggregate changes in the funnel of all ARC colleges and each ARC segment
from the fall 2020 to fall 2024 cycle. The objective of this figure is to compare the last cycle
before widespread test-optional adoption (fall 2020) to the most contemporary cycle (fall 2024).

Figure 3: Percentage Change in Applications, Admissions, and Enrollment Between Fall 2020
and Fall 2024, ARC Colleges and Segments

Between fall 2020 and fall 2024, applications to ARC institutions increase by 38.3%. More
Selective Public ARC colleges experience the largest application growth (42.5%). Admission
offers from ARC institutions also increase, by about 20%. Offers of admission grew by 30% at
Selective Public colleges but declined by 12% at More Selective Private colleges.
In the aggregate, ARC institutions expanded first-year enrollment by 11.2% from fall 2020 to fall
2024. Enrollment grew substantially in all ARC segments except More Selective Private ARC
institutions, where enrollment increased only 0.3% between fall 2020 and fall 2024.

13
Applications

Figure 4 shows how the 38.3% growth in application volume from fall 2020 to fall 2024 across
all ARC institutions breaks down across different student subgroups. Between fall 2020 and fall
2024, applications to ARC colleges grew among all student subgroups. Application growth
between fall 2020 and fall 2024 to ARC colleges is larger among students with high HSGPAs
than students with low HSGPAs, especially in the highest GPA quintile. This likely reflects the
high school grade inflation that took place during the pandemic (Goldhaber and Young, 2024).
Figure 4: Percentage Change in Applications to ARC Colleges Between Fall 2020 and Fall
2024, by Student Characteristics

Changes by disadvantage depend on the metric used. Application growth between fall 2020 and
fall 2024 to ARC colleges is larger among first generation students than non-first generation
students. By contrast, the growth in applications is relatively similar among students from less
challenging and more challenging neighborhoods.

Disaggregated by race/ethnicity, all ARC institutions experience application growth from all
racial/ethnic subgroups between fall 2020 and fall 2024. However, growth is highest among
international students and lowest among White students.

14
Table 1 provides data on year-over-year percentage changes in aggregate applications by
student characteristics (columns 1-4), alongside the aggregate ARC data representing the
overall percentage change from fall 2020 to fall 2024 (column 5), presented in Figure 4.
Table 1: Annual Percentage Changes in Applications to ARC Colleges Between Fall 2020 and
Fall 2024, by Student Characteristics
Percentage Change in Applications
Group Fall 2020 to Fall 2021 to Fall 2022 to Fall 2023 to Fall 2020 to
Fall 2021 Fall 2022 Fall 2023 Fall 2024 Fall 2024
All students 15.2% 7.1% 5.4% 6.3% 38.3%

Gender
Female 17.8% 6.4% 4.3% 5.7% 38.3%
Male 12.1% 7.7% 6.7% 6.8% 37.6%

Race/Ethnicity
Black 14.1% 9.3% 7.1% 11.8% 49.3%
Hispanic/Latinx 11.5% 6.4% 11.3% 11.0% 46.5%
Native 6.0% 12.0% 16.7% 10.6% 53.3%
Native Hawaiian/Pac. Islander 28.3% 4.2% 9.3% 2.5% 49.7%
Two or More Races 17.9% 6.8% 6.6% 6.4% 42.9%
White 15.6% 2.6% 1.1% 1.2% 21.4%
Asian 15.7% 10.0% 9.1% 4.8% 45.6%
International 16.9% 18.2% 6.8% 11.6% 64.6%
Parental education (48
colleges)
First-generation 10.1% 13.9% 13.2% 5.3% 49.7%
Not first-generation 18.2% 3.3% 3.9% 6.7% 35.3%
Parental education unknown 8.5% 16.4% 2.9% 6.1% 37.9%

Income status (16 colleges)

Low-income 19.4% 5.3% 6.1% 6.0% 41.4%
Not low-income 17.3% 5.3% 4.7% 6.1% 37.3%
Income unknown 13.7% 8.2% 5.7% 6.4% 38.4%
Landscape NH challenge
quintile
Lowest 16.5% 5.4% 4.9% 4.7% 34.8%
Second 17.3% 4.2% 5.0% 6.7% 37.0%
Third 13.8% 3.5% 6.0% 6.8% 33.3%
Fourth 9.3% 3.7% 6.2% 10.5% 33.0%
Highest -0.3% 6.1% 9.4% 12.9% 30.7%
Missing 19.5% 20.8% 5.5% 6.5% 62.3%
Recalculated HSGPA quintile
(25 colleges)
Lowest 12.4% -6.4% 5.0% 14.0% 25.9%
Second 10.1% 0.9% 0.2% 8.7% 21.0%
Third 11.7% 7.3% 1.6% 4.8% 27.7%
Fourth 12.8% 15.6% 4.3% 5.7% 43.7%
Highest 22.2% 16.3% 14.8% 7.3% 75.0%
Missing 15.9% 6.8% 4.9% 5.2% 36.7%
Note: Data for students of non-binary/other/unknown gender are omitted due to small sample sizes.

15
Admission Offers

Growth in admission offers parallels the application trends shown in Figure 4. Figure 5 shows
how the 20.3% growth in aggregate admission offers among ARC institutions breaks down
across different student subgroups. Over the fall 2020 and fall 2024 periods, admission offers to
ARC colleges grew among all demographic subgroups except applicants with lower high school
grades.

The growth in admission offers from ARC colleges is larger among underrepresented minority
applicants (and Asian students) than among White applicants, with the largest growth in
admission offers occurring for Black applicants (40%).
From a lens of disadvantage, admission offers increase similarly for students from all five
neighborhood challenge quintiles, and increase more for first generation than for non-first
generation students.

Figure 5: Percentage Change in Admission Offers at ARC Colleges Between Fall 2020 and Fall
2024, by Student Characteristics

16
Table 2 provides data on year-over-year percentage changes in aggregate admission offers
by student characteristics (columns 1-4), alongside the aggregate ARC data representing the
overall percentage change from fall 2020 to fall 2024 (column 5), presented in Figure 5.
Table 2: Annual Percentage Changes in Admission Offers Between Fall 2020 and Fall 2024, by
Student Characteristics
Percentage Change in Admission Offers
Group Fall 2020 to Fall 2021 to Fall 2022 to Fall 2023 to Fall 2020 to
Fall 2021 Fall 2022 Fall 2023 Fall 2024 Fall 2024
All students 6.5% 1.2% 4.3% 7.4% 20.6%

Gender
Female 10.2% 1.8% 2.5% 6.8% 22.9%
Male 1.7% 0.1% 6.6% 7.5% 16.7%

Race/Ethnicity
Black 17.4% 10.2% 3.9% 4.3% 40.1%
Hispanic/Latinx 9.7% 4.0% 7.0% 4.9% 28.1%
Native 6.6% 5.7% 16.8% 0.7% 32.6%
Native Hawaiian/Pac. Islander 22.6% 1.2% 1.9% 2.4% 29.4%
Two or More Races 9.5% 5.4% 7.6% 5.2% 30.7%
White 6.6% -2.7% 0.2% 6.5% 10.7%
Asian 2.7% 5.5% 9.8% 10.4% 31.3%
International -4.1% 3.2% 12.9% 11.2% 24.2%
Parental education (48
colleges)
First-generation 4.8% 6.2% 6.1% 1.0% 19.2%
Not first-generation 5.9% -3.8% 3.2% 7.6% 13.1%
Parental education unknown 9.0% 11.2% 5.7% 10.5% 41.7%

Income status (16 colleges)

Low-income 8.2% -7.1% 5.7% -5.6% 0.3%
Not low-income 2.0% -4.9% 4.2% 1.7% 2.8%
Income unknown 7.7% 3.2% 4.2% 9.4% 26.7%
Landscape NH challenge
quintile
Lowest 6.2% 0.7% 4.0% 7.4% 19.5%
Second 10.2% -0.1% 2.6% 7.6% 21.6%
Third 9.5% 0.5% 3.6% 5.5% 20.4%
Fourth 8.4% 1.3% 2.6% 8.5% 22.1%
Highest 2.6% 6.3% 4.5% 8.8% 24.1%
Missing -1.2% 4.7% 11.1% 6.9% 22.8%
Recalculated HSGPA quintile
(25 colleges)
Lowest -12.6% -31.3% -5.0% 28.7% -26.5%
Second -3.8% -26.8% -3.7% 19.3% -19.1%
Third 1.7% -15.1% -3.3% 5.6% -11.8%
Fourth 5.8% 1.2% -1.2% 2.4% 8.3%
Highest 15.8% 15.0% 11.8% 6.2% 58.1%
Missing 6.8% 3.3% 4.6% 7.7% 24.4%
Note: Data for students of non-binary/other/unknown gender are omitted due to small sample sizes.

17
Enrollment

Figure 6 shows how the 11.2% aggregate enrollment growth among ARC institutions breaks
down across different student subgroups. In general, student subgroups with the largest
increases in admissions offers also experienced the largest growth in enrollment.

Between fall 2020 and fall 2024, enrollment at ARC colleges increases among all student
subgroups except for students with lower high school grades. Enrollment growth is slightly
larger among first generation college students than among non-first generation students, and
larger among low-income students than among non-low-income students. Enrollment growth is
slightly lower among students from the highest challenge neighborhoods than among students
from the lower four neighborhood challenge quintiles. Finally, enrollment growth among URM
students was larger than growth in other student subgroups. White students stand out as
experiencing virtually zero growth in enrollment between fall 2020 and fall 2024.
Figure 6: Percentage Change in Enrollment at ARC Colleges Between Fall 2020 and Fall 2024,
by Student Characteristics

18
Table 3 provides data on year-over-year percentage changes in aggregate enrollment growth
student characteristics (columns 1-4), alongside the aggregate ARC data representing the
overall percentage change from fall 2020 to fall 2024 (column 5), presented in Figure 6.

Table 3: Annual Percentage Changes in Enrollment Between Fall 2020 and Fall 2024, by
Student Characteristics
Percentage Change in First-Year Enrollment
Group Fall 2020 to Fall 2021 to Fall 2022 to Fall 2023 to Fall 2020 to
Fall 2021 Fall 2022 Fall 2023 Fall 2024 Fall 2024
All students 8.0% -1.9% 2.6% 2.3% 11.2%

Gender
Female 10.5% -1.5% 1.1% 2.2% 12.5%
Male 4.9% -2.7% 4.5% 2.0% 8.8%

Race/Ethnicity
Black 14.5% 1.6% 10.4% 2.3% 31.4%
Hispanic/Latinx 11.2% 0.8% 7.4% 2.2% 23.0%
Native 2.7% 5.2% 30.1% -10.5% 25.8%
Native Hawaiian/Pac. Islander 7.8% 9.8% -10.3% 14.7% 21.7%
Two or More Races 9.9% 0.1% 9.6% 0.3% 20.9%
White 6.1% -5.4% -0.7% 1.4% 1.1%
Asian 9.5% 1.8% 4.2% 3.7% 20.4%
International 5.2% 9.0% 0.8% 1.3% 17.1%
Parental education (48
colleges)
First-generation 4.8% 1.9% 7.7% -2.3% 12.4%
Not first-generation 9.8% -4.7% 1.5% 3.8% 10.3%
Parental education unknown 5.1% 3.8% 1.9% 1.5% 12.9%

Income status (16 colleges)

Low-income 12.4% -1.9% 6.9% -7.2% 9.4%
Not low-income 7.8% -2.7% 1.9% -1.8% 4.9%
Income unknown 7.8% -1.7% 2.6% 4.5% 13.6%
Landscape NH challenge
quintile
Lowest 8.0% -2.3% 2.5% 2.8% 11.1%
Second 10.7% -3.4% 1.8% 2.0% 11.0%
Third 11.2% -4.3% 2.4% 2.9% 12.0%
Fourth 6.3% -2.9% 1.8% 5.4% 10.9%
Highest -2.0% -0.6% 6.5% 2.6% 6.5%
Missing 7.2% 8.8% 3.5% -4.7% 15.0%
Recalculated HSGPA quintile
(25 colleges)
Lowest -3.0% -31.5% 16.9% 6.4% -17.5%
Second 4.9% -26.2% 5.5% -0.4% -18.7%
Third 7.6% -16.3% 0.4% -1.2% -10.7%
Fourth 12.4% -0.6% -6.0% 2.8% 7.9%
Highest 18.9% 14.7% 0.8% 9.6% 50.8%
Missing 6.0% 0.1% 4.2% 0.9% 11.6%
Note: Data for students of non-binary/other/unknown gender are omitted due to small sample sizes.

19
Student Body Composition
Racial/Ethnic Composition

Figure 7 shows how the racial/ethnic composition of applicants to ARC college changed
between fall 2018 and fall 2024. Across all ARC colleges, the share of underrepresented
minority (URM) applicants grew steadily from 25.3% in fall 2018 to 28.4% in fall 2024. Annual
changes in racial/ethnic diversity visible in Figure 7 reflect growing diversity of high school
graduates in the U.S. (Lane et al., 2024). Beyond this time trend, there is no abrupt change in
URM student representation between fall 2020 and fall 2021, the first cohort with widespread
test-optional admissions policies, or between fall 2023 and fall 2024, the first cohort with
widespread race-neutral admissions policies.

Figure 7: Racial/Ethnic Composition of Applicants from Fall 2018 to Fall 2024

Figures 8 and 9 show how the racial/ethnic composition of admitted and enrolled students,
respectively, change between fall 2018 and fall 2024 at ARC colleges. The basic patterns for
URM shares are similar to the pattern for applicants; URM students represent roughly one-
quarter of the cohort in the baseline year (fall 2018) and their share increases by about 3
percentage points over the next seven years.

20
Figure 8 shows a 1.3 percentage point increase in the share of admitted URM students between
fall 2020 and fall 2021, which corresponds to the introduction of widespread test-optional
policies. This year-over-year increase is not atypical. In the absence of similar policy shocks,
increases of similar magnitudes occur between fall 2018 and fall 2019, and between fall 2021
and fall 2022.

Figure 8: Racial/Ethnic Composition of Admitted Students from Fall 2018 to Fall 2024

Figure 9 shows the racial/ethnic composition of enrolled students at ARC institutions increases
from 23.8% in fall 2018 to 28.1% in fall 2024. This 4.3 percentage point increase in URM
student representation among ARC college enrollees is nearly identical to the 4.6 percentage
point increase in the share of public high school graduates who are classified as URM students,
which increased from 42.6% to 47.2% over the same time period (Lane et al., 2024).

21
Figure 9: Racial/Ethnic Composition of Enrolled Students from Fall 2018 to Fall 2024

Figure 10 disaggregates the racial/ethnic composition patterns in Figure 9 by institutional

segment to demonstrate more nuanced effects of admissions policy changes in fall 2021 and
fall 2024. More Selective Private ARC colleges experienced an increase in the share of URM
students among first-year enrollees during the first test-optional admissions cycle (a 3.8
percentage point increase from fall 2020 to 2021), while the URM student share of enrollees
remained roughly flat in the other segments.

Following the implementation of race-neutral policies and practices in fall 2024, the share of
URM students at More Selective Private institutions declined by 2.8 percentage points. By
contrast, URM student representation continued to increase gradually through fall 2024 at
institutions in the other three ARC segments. Considerable variation exists across ARC
institutions, with roughly half of institutions experiencing increases and the other half
experiencing decreases in the URM share of enrollees between fall 2023 and fall 2024.

22
Figure 10: Racial/Ethnic Composition of Enrolled Students in Fall 2018 to Fall 2024, by ARC
Segment

Socioeconomic Composition

Figures 11, 12, and 13 show the composition of ARC applicants, admitted students, and
enrollees, respectively, over time and by socioeconomic status. Socioeconomic composition is
described here by neighborhood challenge quintile, a measure of educational opportunity in a
census tract.5 The percentage of applicants from higher challenge neighborhoods (third, fourth,
and highest challenge quintiles) remained relatively steady at 25% from fall 2018 to 2024.
Similarly, the percentage of admitted students from these neighborhoods is constant at 23% for
all seven years. There is a slight increase (<1 percentage point) in the share of applicants and
admitted students from the highest neighborhood challenge quintile from fall 2018 to fall 2024.

5Figures 11, 12, and 13 exclude students missing neighborhood challenge information. Since challenge data are not
available for international students, these figures represent the socioeconomic composition of domestic applicants,
admitted students, and enrollees by neighborhood challenge quintile. See [Link]
[Link]/landscape/[Link] for more information on
neighborhood challenge data.

23
Figure 11: Socioeconomic Composition of Applicants from Fall 2018 to Fall 2024

Figure 12: Socioeconomic Composition of Admitted Students from Fall 2018 to Fall 2024

24
Figure 13 uses the same neighborhood challenge data to show how the socioeconomic
composition of ARC colleges’ first-year enrollees changed between fall 2018 and fall 2024.
Across all ARC colleges, the share of domestic first-year enrollees from neighborhoods in the
three highest challenge quintiles is quite stable at 25%, mirroring the pattern observed among
applicants.

Figure 13: Socioeconomic Composition of Enrolled Students from Fall 2018 to Fall 2024

A more nuanced view emerges when the socioeconomic composition of enrolled students is
examined across the four institutional segments in Figure 14. At the More Selective Private
colleges, the share of students from higher challenge neighborhoods increases from 17.4% to
21.7% between fall 2018 and fall 2024. By contrast, in the other three institutional segments, the
share of enrollees from higher challenge neighborhoods declines slightly over time.

25
Figure 14: Socioeconomic Composition of Enrolled Students from Fall 2018 to Fall 2024, by
ARC Segment

Test Score Choices

Score Disclosure, Withholding, and Absence

Linking ARC institutions’ applicant records to College Board data enables us to distinguish
applicants who disclose scores, who withhold SAT scores, and with no recorded SAT score.
Specifically, we categorize applicants in test-optional cohorts into three groups:

• Score disclosers provided an SAT/ACT score for consideration in admissions.

• Score withholders had an SAT score and withheld it.
• No-SAT applicants had no SAT score and either withheld or had no ACT score.6

6We cannot distinguish ACT withholders from applicants who had no ACT score. As a result, our count of score
withholders is a lower bound on the true number of score withholders; conversely, our count of No-SAT applicants is
an upper bound.

26
No-SAT score applicants and score withholders are indistinguishable to institutions in the sense
that neither group presents a test score in their application, but these two groups are
distinguishable using College Board data.

Figure 15 shows the distribution of applications by score disclosure status in each college
segment and at all ARC institutions, for each cycle from fall 2021 to fall 2024. At all ARC
institutions, test score disclosure rates among applications declined from 52% in 2021 to 48% in
2024 and SAT withholding rates increased from 27% in 2021 to 35% in 2024. The share of
applications in the No-SAT category declined from 22% to 17%. Test score disclosure rates are
higher among applications to More Selective ARC institutions relative to Selective Institutions.

Figure 15: Score Disclosure, Withholding, and Absence Among Fall 2021 to Fall 2024
Applications, by ARC Segment

Score Disclosure Model

In order to understand students’ score choices with greater nuance and while holding other
factors constant, we examine score disclosure and withholding decisions in a statistical model.
Ordinary Least Squares (OLS) regression is the most common statistical method used to
understand what factors influence an outcome. We use OLS to explain what drives the variation

27
in students’ score choices and find that SAT score is the strongest predictor of an applicant’s
decision to disclose their test score. Figure 16 shows the strong correlation between a student’s
SAT score (relative to the college they are applying to) and the probability of disclosing a test
score on their application to an ARC college in the fall 2024 cycle.

Examining combinations of predictors, like SAT score and HSGPA, further reveals that HSGPA
adds minimal incremental predictive strength in determining who will disclose a test score.
Continuing to add predictors that are student demographic characteristics, like race/ethnicity
and parental education, adds minimal to no incremental predictive strength (see Appendix for
regression details).7
Figure 16: Predictors of Test Score Disclosure for Fall 2024

Results from the fully specified OLS model with all predictors are graphically displayed to better
understand any differences in score choices over time or across student subgroups. Figure 17
shows predicted probabilities of test score disclosure for applicants with different test scores

7
The simple (single-variable) correlations with test score disclosure are SAT (0.567), Race/Ethnicity (0.204), HSGPA
(0.136), and Parental Education (0.143). SAT enters the model first because it has the strongest simple correlation.
Although variables could be added in order of their simple correlative strength, it is common to include all academic
achievement variables (e.g., SAT and HSGPA) before adding demographic variables (e.g., Race/Ethnicity and
Parental Education) to better understand the predictive strength of including academics and demographics. Results
are not sensitive to the order in which variables are added to the model. Additional predictors in the final model
include feeder high school status, in-state status, neighborhood and high school challenge, and college-specific
indicators.

28
relative to the college to which they applied, for applicants in the fall 2021 to fall 2024 cohorts.8
For example, a student with a 30th percentile SAT score relative to the college average has a
score that is in the 30th percentile compared to the distribution of applicants’ scores at the ARC
college to which they applied.9 The upward slope of the score disclosure probability curves in
Figure 17 demonstrates the strong relationship between students’ relative test score and the
probability of disclosing a test score in all four test-optional cycles. Applicants with relatively low
test scores have a low predicted probability of disclosing those scores; applicants with relatively
high test scores have a high predicted probability of disclosing those scores, regardless of
application cohort.

Figure 17: Probability of Test Score Disclosure Among ARC College Applications, Fall 2021 to
Fall 2024

In all four years, students with the lowest test scores have a roughly 5% probability of disclosing
scores. By contrast, students with the highest test scores have disclosure probabilities near

8
Probabilities in Figures 18-22 are based on a logistic regression model that also controls for race, in-state status,
first-gen status, feeder school status, HSGPA, neighborhood and high school challenge, and indicators for the
sampled colleges. Very similar patterns emerge in un-modeled data on score disclosure at every ARC institution.
9
The college average SAT score is calculated using data on all SAT and concorded ACT scores from 2018 to 2024
cohorts.

29
100%. Figure 17 reveals that disclosure probabilities changed the most over time for students
with test scores in the middle of the distribution. For example, at the 50th percentile score, the
probability of score disclosure decreases from 57% in fall 2021 to 46% in fall 2024, all else
constant.
The score choice models are useful for examining differences in score disclosure probabilities
for different types of students. Figure 18 reveals differences in predicted score disclosure
behavior among applicants with different high school grades for fall 2024. For example,
applications from students with high school grades of B+ are more likely to include a disclosed
test score compared to applications from students with grades of A+ and A.10 These differences
by HSGPA are starkest in the middle of the test score percentiles. For example, at the 50th SAT
score percentile, students with A+ HSGPAs have a 38% probability of score disclosure,
compared to otherwise similar B+ students, who have a 55% probability of score disclosure.

Figure 18: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by HSGPA

10High school grades below B+ are not displayed because more than 90% of applications to ARC colleges are from
students with HSGPAs of B+ or higher.

30
Figure 19 shows score disclosure probability curves by parental education for the fall 2024
application cohort. Figure 19 demonstrates that first generation and non-first generation
applicants with the same relative test score have virtually identical probabilities of disclosing that
score to ARC colleges.

Figure 19: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by Parental Education

Figure 20 shows score disclosure probability curves by race/ethnicity in the fall 2024 application
cohort. It suggests disclosure behavior does not differ substantially among applicants from most
racial/ethnic subgroups. For example, among applications with low relative scores (e.g., 20th
percentile), all racial/ethnic subgroups share a low probability of disclosing their scores.
However, Asian students are somewhat less likely to disclose at all score levels, relative to other
racial/ethnic subgroups.

31
Figure 20: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by Race/Ethnicity

Figure 21 shows score disclosure probability curves by broad college major category (e.g.,
STEM vs non-STEM) in the fall 2024 application cohort. Whether students intend to major in
STEM fields or non-STEM fields, both groups are equally likely to disclose their test scores
when all other factors are held constant.11

11 Intended college major data come from the SAT/PSAT Student Data Questionnaire.

32
Figure 21: Probability of Test Score Disclosure Among ARC College Applications for Fall 2024,
by College Major

Practically, the logistic regression results presented graphically in Figures 17 to 21 demonstrate

that applicants with the same test score and high school grades make very similar score
choices when applying to ARC colleges. There are no notable differences in score choice
behavior by demographic or academic attributes of applicants. Regression analyses reveal that
test score, relative to the distribution of test scores at the college to which students apply, is the
strongest determinant of a student’s decision to disclose a score.12,13

12 Previously published differences in score disclosure patterns by race, parental education, and income are
attributable to differences in academic achievement among score disclosers and withholders, for which prior analyses
were unable to control (Freeman et al., 2021).
13 The relationship between test scores and disclosure probabilities may not be strictly causal. The statistical models

control for a host of academic and socio-demographic characteristics that simultaneously impact both SAT scores
and score disclosure, however, unobservable characteristics of students may continue to influence the relationships
between SAT scores and disclosure. To address this possibility, McManus et al. (2023) harness data from students
applying to multiple ARC institutions who disclosed scores in a subset of applications. The within-student statistical
models in this research effectively control all observed and unobserved student-level characteristics that are constant
across the colleges to which students applied. They find that a 10-percentile increase in college-specific SAT scores
increases the probability that a student will disclose scores to that college by approximately 5 percentage points.

33
First Year Outcomes

Pre-Collegiate Academic Achievement Differences

In this section, we motivate the analyses of first-year college outcomes by examining pre-
collegiate differences in academic preparation between test score disclosers and non-
disclosers. We also examine how pre-collegiate differences in the academic preparation of
enrollees at ARC institutions have changed in recent years.

Figure 22 shows that enrolled first-year students in fall 2024 who disclosed test scores are more
likely to have higher HSGPAs while scores from non-disclosers are more likely to have lower
HSGPAs. This pattern is apparent in all four ARC institutional segments. For example, at the
More Selective Private colleges, 43% of score disclosers in fall 2024 have “A+” HSGPAs
compared to 34% of withholders. Not surprisingly, HSGPAs are higher among students at More
Selective institutions than at Selective institutions.

Figure 22: Distribution of High School Grades Among Enrolled First-Year Students at ARC
Institutions in Fall 2024, by Score Disclosure Status and ARC Segment

34
Figure 23 demonstrates how high school grades have increased since the pandemic by
comparing the HSGPAs of fall 2024 enrollees with fall 2020 enrollees, the last cohort that
applied to college under broad test-required admission policies. High school grades tend to be
higher by roughly 0.1 HSGPA points among fall 2024 first-year college students relative to fall
2020 in all institutional segments. In 2024, at least 80% of enrollees had HSGPAs of A- or
higher. At More Selective Private colleges, the share of enrollees with “A+” or “A” HSGPAs
increased from 72% of first-year enrollees in fall 2020 to 83% in fall 2024. This grade inflation
during and after the pandemic is consistent with evidence by Sanchez (2023) and Goldhaber
and Young (2023).

Figure 23: Distribution of High School Grades Among Enrolled First-Year Students at ARC
Institutions in Fall 2024 versus Fall 2020, by ARC Segment

Next, we examine patterns in academic preparation as measured by standardized test scores.

Figure 24 shows histograms of SAT scores, separately for score disclosers and withholders, to
reveal bimodal distributions in all four segments. In fall 2024, enrolled first-year students who
disclosed test scores are more likely to have higher scores, by roughly 200 SAT points, on
average, compared to those who withheld scores. This pattern is consistent with prior research
(Howell et al., 2022a) and is evident in all four ARC institutional segments.

35
Figure 24: Distribution of SAT Scores Among Enrolled First-Year Students at ARC Institutions
in Fall 2024, by Score Disclosure Status and ARC Segment

Figure 25 compares the distribution of SAT scores among enrollees in the fall 2024 application cohort
relative to the fall 2020 application cohort. Average test scores decline by approximately 45 SAT
points in all four segments between fall 2020 and fall 2024, but the distribution of test scores in each
segment visibly shifted left in fall 2024 compared to fall 2020, implying an increase in the number of
lower-scoring enrollees under widespread test-optional admissions. Relative to the fall 2020 average,
ARC institutions experienced an increase of nearly 25,000 below-average scoring enrolled first-year
students in fall 2024. This represents a 62% increase in below-average scoring students relative to
fall 2020. All four ARC segments experienced an increase in below-average scoring students relative
to the segment-specific fall 2020 average SAT scores. These segmental increases in below-average
scoring first-year enrollees amount to 47%, 54%, 59% and 85% for More Selective Private, Selective
Private, More Selective Public, and Selective Public segments, respectively.

36
Figure 25: Distribution of SAT Scores Among Enrolled First-Year Students at ARC Institutions
in Fall 2024 versus Fall 2020, by ARC Segment

Changes in academic preparation conveyed in Figure 23 (HSGPA) and Figure 25 (SAT scores) tell
conflicting stories about how the academic preparation of first-year students at consortium institutions
changed in the last five years. Average HSGPAs are increasing, while average test scores are
decreasing. Although first-year college outcomes should be the arbiter of these conflicting data on
preparation, college faculty from over 1,200 postsecondary institutions report incoming college
students since the pandemic are academically weaker than in the past and that faculty are reducing
course content, providing more scaffolding to students, and reducing their grading rigor in response
to this change in academic preparation.14

14 Using a different dataset, other research has similarly demonstrated that HSGPAs (and college GPAs) increased
after the pandemic, while standardized measures like test scores decreased (Westrick et al., 2024). More than three-
quarters of college faculty report post-pandemic students are less prepared with regard to critical thinking, analysis,
and problem-solving skills. Roughly half of faculty report reducing both the rigor of their course and their grading.

37
First-Year College Outcomes
Given changes in the academic readiness of enrollees at ARC institutions since the pandemic,
first-year college student outcomes may also change over time. We examine first-year GPA
(FYGPA), academic risk (defined by earning an FYGPA lower than 3.0), credit accumulation (on
a 30-credit scale), and retention to the second year at the same institution. Table 4 provides
average values of these four college student outcomes over time for all ARC institutions. For the
academic years in which test-optional admissions policies were broadly used to admit students,
average outcomes for students who disclosed, withheld, and had no SAT are also provided to
elucidate the differences across these groups.

Table 4 shows that average FYGPAs across all ARC institutions initially increased with the start
of the pandemic—affecting spring 2020 and the 2020-21 academic year that witnessed either
remote or hybrid classes for many institutions—before declining very slightly. Academic risk of
earning below a 3.0 FYGPA follows this pattern by initially dropping from 24.3% to 15.2% before
increasing to roughly 20%. Credit accumulation and retention rates are quite stable over time,
on average. Both credits earned and retention rates were virtually identical between the pre-
pandemic 2018-19 entering cohort and the students entering during the 2023-24 academic year.
See Appendix Table A4 for segmental views of the average first-year college outcome data in
Table 4.15

Table 4: Average College Outcomes at ARC Institutions, by Academic Year

% of FY Credits
FYGPA FYGPA<3.0 Retention
Enrollees (30 Scale)
2018-19 Overall 3.29 24.3% 29.1 92.8%
2019-20 Overall 3.44 15.2% 29.1 92.2%
2020-21 Overall 3.46 15.2% 28.7 92.2%

2021-22 Overall 3.35 21.5% 28.6 92.4%

nnDisclosers 62.2% 3.40 19.4% 28.7 92.4%
nnWithholders 20.6% 3.28 24.4% 28.6 93.2%
nnNo SAT 17.3% 3.27 25.3% 28.1 90.9%

2022-23 Overall 3.38 20.2% 29.3 92.9%

nnDisclosers 61.3% 3.45 16.8% 29.5 93.5%
nnWithholders 25.2% 3.25 26.0% 29.0 92.3%
nnNo SAT 13.6% 3.28 24.6% 29.1 91.2%

2023-24 Overall 3.41 18.9% 29.1 92.9%

nnDisclosers 58.6% 3.48 16.0% 29.3 93.5%
nnWithholders 28.7% 3.31 22.6% 29.1 92.8%
nnNo SAT 12.7% 3.29 24.3% 28.4 90.4%

15
Segmental views for all first-year college outcome figures below are available in Appendix Figures A3
through A6.

38
First-Year Grades
Figure 26 demonstrates the positive relationship between test scores and average FYGPA. This
positive relationship is well-documented in the research (Westrick et al., 2019; Marini, Westrick,
& Shaw, 2021; Sanchez, 2024) and holds for both score disclosers (left panel) and withholders
(right panel).

Figure 26: Average FYGPA by Test Score Band and Disclosure Status, 2023-24

Even among students with the same HSGPA, Figure 27 demonstrates that test scores have a strong
positive relationship with FYGPA for both score disclosers and withholders.

Figure 27: Average FYGPA by Test Score Band, HSGPA, and Disclosure Status, 2023-24

39
Figure 28 demonstrates how average FYGPA at each ARC institution has changed since the
pandemic by comparing average FYGPA in 2023-24 to average FYGPA in 2018-19. Each dot in
Figure 28 is an ARC institution. Average FYGPAs have increased at all but two ARC
institutions, with the majority of campuses experiencing average FYGPA increases of 0.1 or
more FYGPA points.

Figure 28: Institution-Specific Changes in Average FYGPA, 2023-24 versus 2018-19

First-Year Academic Risk

To focus more explicitly on students at risk of lower performance in their first year of college,
Figure 29 shows the percentage of first-year students with FYGPAs below 3.0 declines rapidly
as test scores increase. This pattern of higher academic risk among those with lower test scores
is as strong for score withholders (right panel) as for disclosers (left panel). Westrick et al.
(2023) demonstrate that students with FYGPAs below 3.0 are at the greatest risk of earning
cumulative college GPAs below 3.0, a threshold often required for continuation in some
programs, to maintain some scholarships, for graduate school admission, and by corporate
recruiters. More than half of the lowest scoring students are at risk of earning a FYGPA below
3.0. While this is true for both score disclosers and withholders, institutions lack visibility into the
risk for score withholders and thus have a reduced capacity to predict which students might
benefit from academic support in their first year of college. Figure 30 shows that the relationship
between test scores and academic risk persists even after controlling for HSGPA.

40
Figure 29: Academic Risk by Test Score Band and Disclosure Status, 2023-24

Figure 30: Academic Risk by Test Score Band, HSGPA, and Disclosure Status, 2023-24

41
Figure 31 demonstrates how academic risk at each ARC institution has changed since the
pandemic by comparing the percentage of enrollees with FYGPAs below 3.0 in 2023-24 to
2018-19. Each dot in Figure 31 is an ARC institution. The percentage of enrollees with FYGPAs
below 3.0 has decreased at all but three ARC institutions, with the majority of campuses
experiencing decreases in academic risk of 3 percentage points or more. The decrease in
academic risk is directly related to higher FYGPAs at nearly every ARC institution.

Figure 31: Institution-Specific Changes in Academic Risk, 2023-24 versus 2018-19

First-Year Credit Accumulation

Figure 32 demonstrates the positive relationship between test scores and average credit
accumulation in the first year of college. This pattern holds for both score disclosers (left panel)
and withholders (right panel). Students with SAT scores below 1200 tend to accumulate fewer
than 30 credits per year and are at risk of not graduating in four years. Even among students with
the same HSGPA, Figure 33 demonstrates that test scores have a strong positive relationship with
first-year credit accumulation for both score disclosers and withholders. Students with the lowest
HSGPAs who earn test scores above 1300 have higher credit accumulation, on average, than
students with the highest HSGPAs and lowest test scores.

42
Figure 32: Average Credits by Test Score Band and Disclosure Status, 2023-24

Note: Credits are normalized to 30 credits per year since many institutions require 120 credits (30 credits per year for
four years) to graduate.

Figure 33: Average Credits by Test Score Band, HSGPA, and Disclosure Status, 2023-24

43
Figure 34 demonstrates how average first-year credit accumulation at each ARC institution has
changed since the pandemic by comparing the average first-year credits earned by enrollees in
2023-24 compared to 2018-19. Each dot in Figure 34 is an ARC institution. Average first-year
credit accumulation has increased slightly in roughly half of institutions and decreased slightly in
half of institutions, with a handful of outliers experiencing more notable changes.

Figure 34: Institution-Specific Changes in Average First-Year Credits, 2023-24 versus 2018-19

Note: Credits are normalized to 30 credits per year.

First-Year Retention
Figure 35 demonstrates the positive relationship between test scores and retention rates into
the second year of college. This pattern holds for both score disclosers (left panel) and
withholders (right panel). The lowest scoring students have first-year retention rates below 85%,
roughly 10 percentage points lower than the retention rates of the highest scoring students.16
Even among students with the same HSGPA, Figure 36 demonstrates that test scores have a
positive relationship with retention for both score disclosers and withholders. Students with the
lowest HSGPAs who earn test scores above 1300 have higher retention rates, on average, than
students with the highest HSGPAs and lowest test scores.

16The relationship between test scores and retention is dampened because some high-achieving students are not
retained when they transfer to more selective postsecondary institutions.

44
Figure 35: Retention by Test Score Band and Disclosure Status, 2023-24

Figure 36: Retention by Test Score Band, HSGPA, and Disclosure Status (2023-24)

45
Figure 37 demonstrates how retention rates into the second year at each ARC institution have
changed since the pandemic by comparing retention rates into the second year among 2023-24
enrollees to 2018-19 enrollees. Each dot in Figure 37 is an ARC institution. Average retention
rates increased at roughly half of institutions and decreased in half of institutions since before
the pandemic, with two notable outliers experiencing decreases in retention rates of 5
percentage points or more.

Figure 37: Institution-Specific Changes in Retention Rates, 2023-24 versus 2018-19

Discussion of College Outcomes

Data on the academic preparation of students after the pandemic suggests readiness has either
increased (according to HSGPA) or decreased (according to test scores). Figure 28, which
shows nearly every ARC institution experienced increases in FYGPA between the 2018-19 and
2023-24 academic years, suggests that we ought to believe the favorable readiness signal of
HSGPA. However, we do not see evidence of similar widespread improvements in credit
accumulation or retention (in Figures 34 and 37, respectively) at ARC institutions over this same
time period. This evidence suggests that college grades may be similarly inflated to high school
grades. Figure 38 examines this question explicitly by showing average FYGPAs changes over
time for students with the same SAT score. Figure 38 clearly demonstrates that students
enrolled at ARC institutions post-pandemic earned demonstrably higher average FYGPAs than
their pre-pandemic peers with the same SAT scores. This evidence of college grade inflation
since the pandemic is evident in all institutional segments (see Appendix Figure A1).

46
Figure 38: Average FYGPA at ARC Institutions by SAT Score Over Time

Outcomes in STEM and Non-STEM College Coursework

The higher first-year college grades shown above result in minimal concerns about negative
effects of the pandemic on college performance. In the 2023-24 academic year, 81.3% of first-
year students at ARC institutions earned FYGPAs above 3.0, which is hardly a cause for alarm,
particularly because that percentage is even higher than in 2018-19 (74.7%). Yet, faculty in
some disciplines have been more vocal about reduced college readiness and first-year success,
and in some rare cases have established policies to combat grade inflation (Butcher, McEwan,
and Weerapana, 2014; Tillinghast et al., 2023; Kuperman et al., 2025). Many of those voices
come from STEM classrooms, so first-year college performance changes may be better
understood through the more nuanced lens of different types of college coursework.

Figure 39 shows the average FYGPA in STEM and non-STEM coursework by SAT score,
regardless of student major, for first-year students admitted under test-optional policies.
Performance in both STEM and non-STEM college courses is positively related to SAT scores.
The visibly higher non-STEM course FYGPAs suggest that STEM course grading practices are
more stringent and/or STEM course content is more difficult compared to non-STEM course
content (Tomkin and West, 2022). On average, first-year performance in STEM courses yields
FYGPAs below 3.0 for students with SAT scores below 1200. By contrast, all but the lowest SAT
scoring students earn FYGPAs above 3.0 in non-STEM coursework.

47
Figure 39 also reveals a steeper positive relationship between STEM course performance and
SAT scores than for non-STEM coursework. The steepness of the STEM line means that
academic risk increases more rapidly as SAT scores fall for students in STEM college
coursework. Despite this increased risk, institutions have less visibility into academic risk of
students in STEM coursework because score disclosure rates fall with SAT scores. For
example, roughly one-quarter of students with SAT scores of 1000 disclose those scores, which
would have signaled a 2.63 FYGPA in STEM courses and a 3.14 FYGPA in non-STEM
coursework. The strength of these results is driven primarily by the strength of the relationship
between SAT Math scores and STEM outcomes, which is shown in Appendix Figure A2
(Westrick, Marini, and Shaw, 2021).

Figure 39: Average FYGPA in STEM and Non-STEM College Coursework in Post-Pandemic
Cohorts, by SAT Score

Note: Data represent three cohorts of students who first entered ARC institutions in the fall of 2021, 2022, and 2023.

48
Conclusion
This research provides the most extensive evidence to date on how selective colleges have
adapted to and been shaped by two major recent shifts in admissions policy—widespread test-
optional policies beginning after the pandemic and the advent of widespread race-neutral
admissions following summer 2023 Supreme Court rulings. Across more than 60 institutions
participating in the Admissions Research Consortium, findings reveal substantial growth in
applications to consortium institutions with notable increases in admissions offers and
enrollments at all but the More Selective Private institutions.

Our analyses show that the demographic composition of consortium members’ student bodies
has changed minimally. Test-optional policies are associated with modest gains in
underrepresented minority (URM) student enrollment at More Selective Private colleges in fall
2021, but race-neutral policies effective for fall 2024 partially reversed these gains. Overall, the
socioeconomic composition of enrolled students remains unchanged over time.

Student SAT scores, relative to other applicants at the same institution, remain the primary
driver of test score disclosure and withholding choices. Other student-level characteristics like
race/ethnicity, parental education, or college major are relatively unimportant in explaining
students’ score choices. SAT scores, whether disclosed or withheld, remain highly predictive of
student outcomes, like first-year GPA (FYGPA), credit accumulation, and retention. Such
relationships persist after controlling for high school GPA. This connection between test scores
and first-year outcomes is especially pronounced in STEM coursework, in which students with
lower test scores face higher risks of performing poorly. These same lower scoring students are
also less visible to the institutions because of their tendency to withhold scores.

While average FYGPAs have risen across ARC institutions, this growth in performance likely
reflects college grade inflation rather than improvements in academic preparedness. Among
students with the same SAT scores, FYGPAs have increased by as much as 0.30 GPA points,
relative to pre-pandemic FYGPAs. By contrast, first-year credits earned and retention have
remained flat since the pandemic, despite declines in entering student academic preparation.
College faculty have voiced concerns over declines in students’ academic preparation and are
reporting adjustments in course structure, content, and grading in response to declines in
student preparation. Future research will examine retention to third and fourth year of college as
well as bachelor’s degree completion.

49
Appendix
Table A1 compares ARC institutions to other four-year non-ARC institutions in the same
segment for sample 1 (e.g., the application/admissions/enrollment sample). Table A2
makes the same comparisons between ARC and non-ARC institutions in sample 2 (e.g.,
the first-year college outcomes sample).

Table A1: Attributes of ARC and Non-ARC Institutions by Segment, Sample 1

More Selective Selective Private More Selective Selective Public
Private Colleges Colleges Public Colleges Coleges
Non- Non- Non- Non-
ARC ARC ARC ARC ARC ARC ARC ARC
Number of Institutions 24 52 18 1,015 13 73 8 450
First-Year Enrollment 1,496 799 1,049 387 6,279 2,784 6,115 1,709
Admit Rate 11% 13% 45% 74% 46% 43% 70% 85%
Yield Rate 49% 55% 22% 32% 28% 33% 22% 25%
First-Year Pell Share 17% 18% 15% 38% 24% 44% 20% 41%
Percent In-State 18% 23% 35% 60% 79% 74% 46% 81%
Tuition and Fees $64,948 $51,829 $58,018 $32,642 $12,940 $10,035 $13,001 $10,462
Graduation Rate 93% 88% 84% 56% 83% 65% 74% 51%

First-Year Racial/Ethnic Composition

Native American 0% 0% 0% 1% 0% 0% 0% 1%
Asian American 18% 14% 8% 3% 18% 11% 6% 5%
Black/African American 7% 9% 4% 12% 8% 18% 8% 15%
Hispanic 13% 11% 14% 16% 17% 25% 11% 18%
Native Hawaiian 0% 0% 0% 0% 0% 0% 0% 0%
White 40% 41% 59% 55% 44% 35% 66% 51%
Two Or More Races 7% 7% 6% 4% 5% 4% 5% 5%
Unknown 2% 4% 1% 4% 2% 2% 1% 3%
International 13% 13% 8% 5% 5% 4% 2% 2%

50
Table A2: Attributes of ARC and Non-ARC Institutions by Segment, Sample 2
More Selective Selective Private More Selective Selective Public
Private Colleges Colleges Public Colleges Colleges
Non- Non- Non- Non-
ARC ARC ARC ARC ARC ARC ARC ARC
Number of Institutions 23 53 16 1,017 9 77 8 450
First-Year Enrollment 1,531 797 1,237 385 5,596 3,046 5,995 1,711
Admit Rate 11% 13% 42% 74% 47% 43% 67% 85%
Yield Rate 48% 56% 22% 32% 26% 33% 23% 25%
First-Year Pell Share 17% 18% 14% 38% 24% 43% 21% 41%
Percent In-State 18% 23% 34% 60% 77% 74% 51% 81%
Tuition and Fees $64,891 $52,101 $58,553 $32,684 $13,047 $10,181 $13,453 $10,453
Graduation Rate 93% 88% 85% 56% 82% 66% 76% 51%

First-Year Racial/Ethnic Composition

Native American 0% 0% 0% 1% 0% 0% 0% 1%
Asian American 18% 14% 8% 3% 20% 11% 10% 5%
Black/African American 7% 9% 4% 12% 9% 18% 8% 15%
Hispanic 13% 11% 13% 16% 14% 25% 9% 18%
Native Hawaiian 0% 0% 0% 0% 0% 0% 0% 0%
White 40% 41% 59% 55% 44% 35% 63% 51%
Two Or More Races 7% 7% 5% 4% 5% 4% 5% 5%
Unknown 2% 4% 1% 4% 3% 2% 2% 3%
International 13% 13% 8% 5% 5% 4% 3% 2%

51
Table A3: Average First-Year College Outcomes by Score Disclosure, Withholding
and Absence
SAT/ACT SAT
No SAT
Disclosers Withholders
Average FYGPA
All ARC Institutions 3.48 3.31 3.29
More Selective Privates 3.68 3.46 3.45
More Selective Publics 3.50 3.25 3.27
Selective Privates 3.54 3.35 3.32
Selective Publics 3.32 3.22 3.08

Percent with FYGPA<3.0

All ARC Institutions 16.0% 22.6% 24.3%
More Selective Privates 5.5% 13.8% 15.3%
More Selective Publics 14.2% 26.0% 25.4%
Selective Privates 12.2% 20.4% 22.9%
Selective Publics 24.9% 28.4% 35.5%

Average FY Credits
All ARC Institutions 29.3 29.1 28.4
More Selective Privates 31.2 30.5 30.3
More Selective Publics 28.9 28.2 27.7
Selective Privates 30.4 29.3 28.8
Selective Publics 28.2 28.6 26.8

Average FY Retention
All ARC Institutions 93.5% 92.8% 90.4%
More Selective Privates 96.4% 96.2% 95.0%
More Selective Publics 94.3% 92.3% 90.5%
Selective Privates 93.2% 91.4% 89.9%
Selective Publics 91.0% 90.6% 83.0%

52
Table A4: Average First-Year College Outcomes, by Academic Year and ARC Segment
More Selective Private Selective Private
% of % of
enrollees FYGPA FYGPA<3.0 FY Credits (30 Scale) Retention enrollees FYGPA FYGPA<3.0 FY Credits (30 Scale) Retention
2018-19 Overall 3.45 14.0% 30.8 96.0% 3.36 20.1% 29.7 92.3%
2019-20 Overall 3.57 7.9% 30.9 93.5% 3.47 13.4% 29.8 90.7%
2020-21 Overall 3.60 7.6% 30.6 95.3% 3.41 17.2% 29.2 91.4%
2021-22 Overall 3.53 11.3% 30.4 95.9% 3.39 19.0% 29.2 91.4%
nnDisclosers 54.5% 3.63 7.0% 30.7 96.2% 56.1% 3.47 15.3% 29.7 92.2%
nnWithholders 26.6% 3.42 16.3% 30.1 96.3% 21.4% 3.29 23.4% 28.7 90.9%
nnNo SAT 18.9% 3.42 16.6% 29.8 94.6% 22.5% 3.28 24.3% 28.5 89.9%
2022-23 Overall 3.54 10.9% 32.6 96.0% 3.42 17.1% 29.7 91.8%
nnDisclosers 55.2% 3.65 6.5% 32.8 96.2% 53.7% 3.52 12.5% 30.2 92.5%
nnWithholders 28.6% 3.40 16.8% 32.0 95.8% 27.7% 3.30 23.1% 29.2 91.4%
nnNo SAT 16.3% 3.44 15.5% 32.8 95.9% 18.6% 3.32 21.5% 29.1 90.3%
2023-24 Overall 3.58 9.7% 30.8 96.1% 3.44 16.7% 29.7 92.1%
nnDisclosers 52.3% 3.68 5.5% 31.2 96.4% 50.8% 3.54 12.2% 30.4 93.2%
nnWithholders 32.4% 3.46 13.8% 30.5 96.2% 31.4% 3.35 20.4% 29.3 91.4%
nnNo SAT 15.3% 3.45 15.3% 30.3 95.0% 17.8% 3.32 22.9% 28.8 89.9%
More Selective Public Selective Public
% of % of
enrollees FYGPA FYGPA<3.0 FY Credits (30 Scale) Retention enrollees FYGPA FYGPA<3.0 FY Credits (30 Scale) Retention
2018-19 Overall 3.30 23.4% 28.7 93.3% 3.12 36.1% 27.8 89.3%
2019-20 Overall 3.45 14.8% 28.6 93.3% 3.32 22.4% 28.1 90.3%
2020-21 Overall 3.49 12.9% 28.4 92.9% 3.31 23.5% 27.4 88.9%
2021-22 Overall 3.34 21.5% 27.9 92.8% 3.19 31.0% 27.7 89.0%
nnDisclosers 59.7% 3.40 18.5% 28.2 93.4% 73.9% 3.23 29.4% 27.8 89.1%
nnWithholders 21.8% 3.24 26.2% 27.6 92.8% 14.0% 3.12 34.8% 28.1 90.9%
nnNo SAT 18.6% 3.26 25.6% 27.5 91.1% 12.0% 3.08 37.2% 26.8 86.3%
2022-23 Overall 3.38 19.8% 28.5 93.6% 3.23 29.0% 27.7 90.0%
nnDisclosers 62.3% 3.47 15.1% 28.9 94.8% 68.0% 3.29 26.3% 28.0 90.7%
nnWithholders 24.0% 3.20 28.8% 28.0 92.0% 22.8% 3.14 33.3% 27.4 89.6%
nnNo SAT 13.7% 3.25 25.4% 27.9 90.9% 9.3% 3.05 38.5% 26.1 84.8%
2023-24 Overall 3.40 19.0% 28.6 93.3% 3.27 26.7% 28.2 90.3%
nnDisclosers 58.0% 3.50 14.2% 28.9 94.3% 67.1% 3.32 24.9% 28.2 91.0%
nnWithholders 29.7% 3.25 26.0% 28.2 92.3% 23.8% 3.22 28.4% 28.6 90.6%
nnNo SAT 12.3% 3.27 25.4% 27.7 90.5% 9.0% 3.08 35.5% 26.8 83.0%

53
Figure A1: Average FYGPA at ARC Institutions by SAT Score Over Time, by ARC Segment

54
Figure A2: Average FYGPA in STEM and Non-STEM College Coursework in Post-Pandemic
Cohorts, by SAT Math Score

55
Appendix Figures with First-Year College Outcomes by Institutional Segment

Figure A3: Average FYGPA by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment

Figure A4: Academic Risk by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment

56
Figure A5: Average Credits by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment

Figure A6: First-Year Retention by Test Score Band and Disclosure Status, 2023-24 by ARC
Segment

57
Appendix on Test Score Disclosure Regression Details
In order to estimate the how the probability of SAT score disclosure (𝑝𝐷𝑖𝑠𝑐𝑙𝑜𝑠𝑒 ) varies with a
student’s SAT score, we fit the ARC data with the logistic regression model expressed through
Equation (1). In this model, the variables are defined as follows.
• SATScore is a student’s actual SAT score inclusive of disclosed and withheld scores.
Disclosed scores are sourced from ARC colleges, while withheld scores are the highest
combination of SAT section scores from across all of a student’s SAT scores in College
Board administrative data.
• Race is a vector of indicator variables expressing the student race/ethnicity provided by the
college.
• ParentalEd is a vector of indicator variables expressing the student’s parental education
from the College Board’s Student Data Questionnaire (SDQ).
• InState is an indicator for whether the student resides in the same state as the college to
which they applied.
• HSGPA is the student’s self-reported high school GPA from the College Board’s Student
Data Questionnaire (SDQ) on a 0-4.33 scale.
• HSChallenge and NHChallenge express the student’s high school and neighborhood
challenge on a 1-100 scale, where higher challenge levels indicate more disadvantaged
neighborhoods and high schools.
• Feeder is an indicator variable identifying students that attend a high school that sent 30 or
more applications to the college between 2018 and 2020.
• College is a vector of college fixed effects to capture differences across ARC institutions
that are constant for all students.
In equation (1), parameter 𝜏 identifies the interaction of the Race indicator variables and the
student’s SAT score. This interaction allows for different relationships between the student’s
SAT score and disclosure probability for different subgroups of students, defined by race. In
models where we identify the relationships between HSGPA or Parental Education and SAT
score disclosure, parameter 𝜏 expresses the interaction of these variables and SAT score.
To construct the fitted score disclosure curves depicted in Figures 13-16, we hold all variables
constant at their sample means and use parameters 𝛽, 𝛾, and 𝜏 to demonstrate how SAT
disclosure probabilities change with SAT scores for different subgroups of students.
Equation (1):
𝑝𝐷𝑖𝑠𝑐𝑙𝑜𝑠𝑒
𝑙𝑛 ( ) = 𝛽0 + 𝛽𝑆𝐴𝑇𝑆𝑐𝑜𝑟𝑒𝑖 + 𝛾𝑅𝑎𝑐𝑒𝑖 + 𝜏𝑆𝐴𝑇𝑆𝑐𝑜𝑟𝑒𝑖 ∗ 𝑅𝑎𝑐𝑒𝑖 + 𝛿𝑃𝑎𝑟𝑒𝑛𝑡𝑎𝑙𝐸𝑑𝑖 +
1 − 𝑝𝐷𝑖𝑠𝑐𝑙𝑜𝑠𝑒

𝜁𝐼𝑛𝑆𝑡𝑎𝑡𝑒𝑖 + 𝜙𝐻𝑆𝐺𝑃𝐴𝑖 + 𝜋𝐻𝑆𝐶ℎ𝑎𝑙𝑙𝑒𝑛𝑔𝑒𝑖 + 𝜃𝑁𝐻𝐶ℎ𝑎𝑙𝑙𝑒𝑛𝑔𝑒𝑖 + 𝜕𝐹𝑒𝑒𝑑𝑒𝑟𝑖 + 𝐶𝑜𝑙𝑙𝑒𝑔𝑒𝑖 + 𝜖𝑖

58
References
Butcher, K., P. McEwan, and A. Weerapana. 2014. The Effects of an Anti-Grade-Inflation Policy
at Wellesley College. Journal of Economic Perspectives, 28 (3): 189–204. Retrieved from
[Link]

Edwards, A., P. Goyer, J. Howell, M. Hurwitz, S. Imlay, and G. Perfetto. 2023. College Outcomes
Following Pandemic-Induced Changes in College Admissions Policies: Focus on 2021-22 First-
Year Outcomes. College Board: New York, NY. July 2022. Retrieved from
[Link]

Goldhaber, D. and M. Young. 2023. Course Grades as a Signal of Student Achievement:

Evidence on Grade Inflation Before and After COVID-19. CALDER Research Brief, No. 35.
Retrieved from [Link]
[Link].

Goyer, P., J. Howell, M. Hurwitz, S. Imlay, and J. Ma. 2023. Updated Evidence on Changes in
College Applications, Admission, and Enrollments: Focus on the Fall 2022 Admissions Cycle.
College Board: New York, NY. July 2022. Retrieved from
[Link]

Howell, J., M. Hurwitz, S. Imlay, G. Perfetto. 2022. New Evidence on Recent Changes in
College Applications, Admissions, and Enrollments: Focus on the Fall 2021 Admissions Cycle.
College Board: New York, NY. July 2022. Retrieved from
[Link]

Kim, B., Freeman, M, Kajikawa, T., Karimi, H., and Magouirk, P. (2022). First-year applications
per applicant: Patterns of high-volume application activity at Common App. Common Application
Research Brief. Retrieved from [Link]

Kuperman, V., E. Geva, V. Taler, and K. Theriault. 2025. Recover from University Grade Inflation
After the COVID-19 Pandemic Varies by Faculty. Studies in Higher Education, 1-16. Retrieved
from [Link]

Lane, P., C. Falkenstern, and P. Bransberger. 2024. Knocking at the College Door: Projections
of High School Graduates. Western Interstate Commission for Higher Education. Retrieved from
[Link]

Marini, J., P. Westrick, and E. Shaw. 2021. Examining the Stability of SAT Predictive
Relationships Across Cohorts and Over Time. College Board: New York, NY. Retrieved from
[Link]
[Link].

McManus, B., M. Hurwitz, and J. Howell. 2023. Strategic Disclosure of Test Scores: Evidence
from U.S. College Admissions. EdWorkingPaper: 23-843. Retrieved from Annenberg Institute at
Brown University: [Link]

59
Sanchez, E. and R. Moore. 2022. Grade Inflation Continues to Grow in the Past Decade. ACT
Research Report. Retrieved from [Link]

Sanchez, E. 2023. Evidence of Grade Inflation Since 2010 in High School English,
Mathematics, Social Studies, and Science Courses. ACT Research Report. Retrieved from
[Link]
[Link].

Sanchez, E. 2024. Changes in the Predictive Validity of High School Grade Point Average and
ACT Composite Score After the COVID-19 Pandemic. ACT Research Report. Retrieved from
[Link]
[Link].

Tillinghast, J., J. Mjelde, and A. Yeritsyan. 2023. COVID-19 and Grade Inflation: Analysis of
Undergraduate GPAs During the Pandemic. SAGE Open, 13 (4). Retrieved from
[Link]

Tomkin, J. and M. West. 2022. STEM Courses are Harder: Evaluating Inter-Course Grading
Disparities with a Calibrated GPA Model. International Journal of STEM Education, 9 (27).
Retrieved from [Link]

Westrick, P., J. Marini, L. Young, H. Ng, D. Shmueli, and E. Shaw. 2019. Validity of the SAT for
Predicting First-Year Grades and Retention to the Second Year. College Board: New York, NH.
Retrieved from [Link]

Westrick, P., J. Marini, L. Young, H. Ng, and E. Shaw. 2023. The Consequences of a Low First-
Year Grade Point Average on Later College Outcomes. College Board: New York, NY.
Retrieved from
[Link]
st-Year%20Grade%20Point%20Average%20on%20Later%20College%[Link].

60
About the College Board
The College Board is a mission-driven not-for-profit organization that connects students
to college success and opportunity. Founded in 1900, the College Board was created to
expand access to higher education. Today, the membership association is made up of
over 6,000 of the world’s leading educational institutions and is dedicated to promoting
excellence and equity in education. Each year, the College Board helps more than
seven million students prepare for a successful transition to college through programs
and services in college readiness and college success — including the SAT® and the
Advanced Placement Program®. The organization also serves the education
community through research and advocacy on behalf of students, educators and
schools. For further information, visit [Link].

College Board Research

The Research Department generates data and evidence on the impact of educational
programs, assessments, and initiatives on students and various education stakeholders.
For further information and publications, visit [Link]

61